In sexually reproducing organisms, the halving of the DNA content of a diploid germ line cell during the meiotic cell cycle allows the production of haploid gametes. During this process, recombination plays a dual role: it shuffles information along the lengths of homologous chromosomes, creating genetic diversity that is transmitted to progeny, and it ensures the proper segregation of homologs to opposite poles during the first of the two meiotic divisions. About a century ago, de Vries-predicted that exchanges take place between homologous maternal and paternal chromosomes during hereditary transmission. Soon thereafter in 1905, Bateson-discovered partial linkage between the petal color and pollen shape characters in Sweet Pea. These and successive discoveries in the emerging field of recombination led to the notion of genetic distances, as measured by the frequency of exchange (crossing-over) between linked markers, and to the development of linkage maps. In 1913, Sturtevant wrote “Of course, there is no knowing whether or not these distances as drawn represent the actual relative spacial distances apart of the factors”. Since the advent of the molecular era, this prescient insight has been extensively verified by quantitative comparisons of genetic and physical distances. For all organisms, including the yeast Saccharomyces cerevisiae, Arabidopsis thaliana, Drosophila melanogaster, Mus musculus, and man, meiotic recombination rates (expressed as cM/kb) vary by several orders of magnitude along chromosomes. Recent studies of S. cerevisiae strongly suggest that most of this variation is related to the frequencies of initiating events (Baudat and Nicolas, 1997), but why it is relatively frequent at some loci (hotspots) and relatively infrequent at others (coldspots) remain unexplained.
In S. cerevisiae, meiotic recombination results from the formation and repair of programmed DNA double-strand breaks (DSBs) (for review, see for example: Smith and Nicolas, 1998). Numerous studies have shown that natural DSB sites are not evenly distributed and that cleavage frequencies vary 10–100-fold from site to site (Baudat and Nicolas, 1997; Gerton et al., 2000). The factors that determine whether a specific region or site is prone to DSB formation (and hence, recombination) are not completely understood, but they are known to act both locally and globally. Locally, gene organization and chromatin structure appear to be of paramount, and related, importance. Typically, most natural DSB sites are in promoter-containing regions (Baudat and Nicolas, 1997). At the HIS4 locus, two types of recombination hotspots have been distinguished, α (transcription factors-dependent but not transcription-dependent) and β (transcription factor independent) hotspots (reviewed by Petes, 2001). More notably, all known DSB sites are located in regions that are sensitive to DNase I or micrococcal nuclease (MNase I) in both mitotic and meiotic cells, suggesting that an open chromatin configuration is necessary for cleavage (Ohta et al., 1994; Wu and Lichten, 1994). However, local chromatin accessibility cannot be the sole arbiter of DSB site selectivity, because not all nuclease-hypersensitive sites are DSB sites.
Global determinants also control the distribution of DSBs. Both the fine mapping of DSB sites on yeast chromosome III and the genome-wide mapping of DSB sites have confirmed the existence of large subchromosomal domains hot or cold for DSB formation (Baudat and Nicolas, 1997; Gerton et al., 2000). The molecular basis of these DSB-proficient or -refractory domains has not been elucidated, but the finding that a recombination-proficient reporter inserted at various sites along chromosome III adopts local properties with respect to DNaseI sensitivity and frequencies of DSB formation and recombination demonstrates that domain-level controls are superimposed on local determinants (Borde et al., 1999). That is, a hot region can be made cold, but thus far the converse has not been observed: cold regions typically remain cold.
A consideration of the chromosomal variation in DSB frequencies must also take into account the influence of trans-acting factors. A large number of genes are required for DSB formation, including SPO11, MEI4, MER1, MER2/REC107, MRE2/NAM8, MRE11, RAD50, REC102, REC103/SKI8, REC104, REC114 and XRS2, but in most cases, their molecular roles are unknown. Null mutants for all of the above genes fail to carry out meiotic recombination and produce inviable spores. Three other meiosis-specific genes are required for full levels of DSBs: MEK1/MRE4 encodes a kinase that regulates the activities of the RED1 and HOP1 products, which are structural components of meiotic chromosomes. SPO11 encodes a protein that shares sequence similarity with the smaller subunit (Top6A) of the type II topoisomerase of the archaebacterium Suffobolus shibatae (Bergerat et al., 1997). Spo11 remains covalently linked to the 5′-strand termini of DSB fragments in mutants (e.g. rad50S) that are defective for the 5′ to 3′ nucleolytic processing of DSB ends that normally precedes repair (Keeney et al., 1997), indicating that it is the catalytic component of the meiotic DSB cleavage activity. These and further molecular and genetic studies in fungi and higher eukaryotes have demonstrated that Spo11 orthologs are likely universally required for meiotic recombination, strongly suggesting that DSBs initiate meiotic recombination in most if not all eukaryotes. The use of site-directed mutagenesis to identify regions of Spo11 that contribute to strand cleavage and DNA binding has demonstrated the functional significance of structural motifs conserved throughout the Spo11/Top6A family (Bergerat et al., 1997; Diaz et al., 2002). Interestingly, variations in the level and distribution of DSBs at the his4::LEU2 hotspot in some spo11 mutants suggests that Spo11 is not only involved in the cleavage activity but also contributes to the choice of site for DSB formation, at least locally (Diaz et al., 2002).